Geriatric Depression Scale (GDS)

Geriatric Depression Scale – 30 item
StudyIdentification toolComparator/ casenessPopulationResults
Consultation
Blank2004

Quality assessed: +
GDS – 30Diagnostic Interview ScheduleN = 360, participants were recruited from primary care (N=125), general hospitals (N=150) and nursing home (N=85) settings (analysis presented separately for each group). All participants were aged >60 years

Mean age – 77 years

Male = 37%

Prevalence of major depression – 9%

Prevalence of any depression – 16%

Prevalence of major depression in primary care – 11%

Prevalence of major depression in hospital – 8%

Prevalence of major depression in nursing homes – 9%
Major depression

Primary care sample

GDS-30
Cut-off ≥10
Sensitivity – 79% (50-94) Specificity - 67% (63-69)

AUC – 0.87 (0.77-0.97)

Cut-off ≥17 – recommended
Sensitivity – 79% (51-94)
Specificity - 87% (84-89)

Nursing Home sample
GDS-30
Cut-off ≥10
Sensitivity – 86% (44-99)
Specificity - 72% (68-73)

AUC – 0.88 (0.74- 1.02)

Cut-off ≥13 – recommended
Sensitivity – 86% (44-99)
Specificity - 85% (81-86)

Hospital sample

GDS-30
Cut-off ≥10
Sensitivity – 83% (52-97)
Specificity - 78% (75-79)

AUC – 0.90 (0.81- 1.00)

Cut-off ≥15 – recommended
Sensitivity – 83% (54-97)
Specificity - 93% (90-94)
Burke 1992

Quality assessed: +
GDS-30DSM-III-RN = 67 cognitively intact outpatients

Mean age = 77.2 (SD 6.5)

Male = 34%

Prevalence of depression16/67
Depression

Cut-off ≥ 11
Sensitivity – 81%
Specificity – 61%

Cut-off ≥ 14
Sensitivity – 44%
Specificity – 75%

Cut-off ≥ 17
Sensitivity – 31%
Specificity – 94%
Evans 1993

Quality assessed: +
GDS-30Geriatric Mental State (GMS)N = 408, older adults attending primary care, London. N = 144 randomly selected for analysis of GDS

Mean age of total sample – 73 years (SD – 8.4)

Male – 38% of total sample

Prevalence of depression59/144
Depression

GDS
Sensitivity – 0.8475
Specificity – 0.7176
Fernandez-San Martin 2002

Quality assessed: +
GDS-30DSM-IVN=192 age >65 years, gender: 70 males, 122 females

Primary care, Spain

Prevalence of depression - 60/192 (mainly psychotic depression)
Depression

Cut-off ≥11
Sensitivity = 0.817
Specificity = 0.68
Jongenelis 2005

Quality assessed: +
GDS-30DSM-IVN= 333, age = 79 years, gender: 104 males, 229 females

Nursing home, Netherlands

Prevalence of depression -: 74/333
Any depression

Cut-off 11
Sensitivity - 0.85
Specificity - 0.69
Koenig 1992A

Quality assessed: +
GDS-30DSM-III-RN = 109 medically ill hospitalized patients

Mean age = 74 (S.D. 4.1)

100% men

Mean MMSE score = 25.7 (S.D. 3.3)

US, Durham

Prevalence of depression11/109
Major depression

Cut-off ≥ 11 - GDS
Sensitivity – 82%
Specificity - 76%
PPV – 27%
NPV – 97%
Laprise 1998

Quality assessed: +
GDS-30DSM-III-RN=66, Nursing home residents, Canada (French)

Mean age = 78 years,

gender: 31 males, 35 females

Prevalence of depression27/66
Depression

Cut-off 10-GDS
Sensitivity = 0.92
Specificity = 0.513
Lynes 1997

Quality assessed: +
GDS – 30DSM-III-RN = 130 older adults attending primary care.

Mean age – 71.0 years (SD – 6.8 years)

Male – 41.5%

Prevalence of major depression – 14/130

Prevalence of any depression – 24/130
Major depression

Cut-off 10 GDS-30
Sensitivity = 100%
Specificity = 84%

AUC – 0.936 (0.031)
Magni 1986

Quality assessed: ++
GDS-30DSM-IIIN = 220, Consecutive admissions to general medical ward, Italy

Mean age = 76 years,

Gender: 111 males, 109 females

Prevalence of depression (MDD and dysthymia)67/220
MDD only – 18/220
Depression

Cut-off 11 -GDS
Sensitivity = 0.86
Specificity = 0.74

Cut-off 14 - GDS
Sensitivity = 0.65
Specificity = 0.91
McGivney 1994

Quality assessed: +
GDS - 30DSM-III-RN = 66 new admissions to two nursing homes.

Mean age - 83 years (SD=4)

Male – 29%

Prevalence of major depression::- 6/66

Prevalence of any depression::- 30/66
Any depression

Cut-off≥ 10 – GDS-30
Sensitivity – 63%
Specificity – 83%
Nam Bae 2004

Quality assessed: ++
GDS – Korean version (GDS-K)DSM-III-RN = 154 (91.1% of eligible participants)

Consecutively registered elderly psychiatric patients aged 55+ who visited the Geriatric Psychiatry Clinic in Seoul. People with dementia or any form of cognitive impairment were excluded from the study.

Mean age = 66 years (SD = 6.48yrs)

Male – 35%

Prevalence of depression 62/154
Major depression

GDS-K
Optimal cut-off ≥ 16
Sensitivity = 0.9032
Specificity = 0.7174

Optimal cut-off ≥ 18 (indicated by ROC curve)
Sensitivity = 0.8387
Specificity = 0.8152
Neal 1994

Quality assessed: +
GDS-30GMS-AGECATN = 45 older adults attending medical outpatient clinics in three UK hospitals.

Mean age – 77.2

Male – 38%

Prevalence of depression:- 10/45 (22%)
Depression

Cut-off ≥ 9 – GDS-30
Sensitivity – 0.63
Specificity – 0.80
PPV – 0.92
NPV – 0.38

Cut-off ≥10 – GDS-30
Sensitivity – 0.74
Specificity – 0.80
PPV – 0.93
NPV – 0.47

Cut-off ≥11 – GDS-30
Sensitivity – 0.73
Specificity – 0.80
PPV – 0.94
NPV – 0.57

Cut-off ≥12 – GDS-30
Sensitivity – 0.83
Specificity – 0.80
PPV – 0.94
NPV – 0.57

Cut-off ≥13 – GDS-30
Sensitivity – 0.83
Specificity – 0.70
PPV – 0.91
NPV – 0.54

Cut-off ≥14 – GDS-30
Sensitivity – 0.83
Specificity – 0.60
PPV – 0.88
NPV – 0
Pomeroy 2001

Quality assessed: +
GDS - 30ICD-10N = 87 patients over the age of 60 admitted to medical rehabilitation wards or attending day rehabilitation facilities; 40% male, mean age 78.4 (SD – 7.7 yrs)

Prevalence of depression17/87
Depressive episode

GDS-30
Optimal cut-off ≥ 11
Sensitivity – 100%
Specificity – 62.9%
AUC – 0.85 (0.77, 0.94)
PPV – 39.5%
NPV – 100%
Robison 2002

Quality assessed: +
GDS-30CIDIN=303 Age = 61 years gender: 88 males, 215 females

Primary care, Hispanic population in US

Prevalence: 67/303
Sensitivity = 0.81
Specificity = 0.65
Snowdon 1990

Quality assessed: +/- unable to assess due to lack of information.
GDS-30DSM-IIIN = 69 residents in old age hostels or nursing homes

Mean age – not reported

Male – percentage not reported

Prevalence of major depression – 12/69

Prevalence of any depression – 15/69
Any depression

All participants
Cut-off ≥11 – GDS-30
Sensitivity – 93%
Specificity – 83%

Cut-off ≥14 GDS-30
Sensitivity – 60%
Specificity – 94%

Nursing home participants only
Cut-off ≥11 – GDS-30
Sensitivity – 100%
Specificity – 66%

Cut-off ≥14 GDS-30
Sensitivity – 71%
Specificity – 92%
Van Marwijk 1995

Quality assessed: +
GDS – 30 itemDSM-IIIN=586 age = 65-94 years, gender: 237 males, 349 females

Older people in primary care, Netherlands

Prevalence of depression - 33/586
Any depression

Cut-off 10 – GDS-30
Sensitivity = 0.55
Specificity = 0.86
Vargas 2007

Quality assessed: +
GDS-30DSM-IVN=484 age = 70 years, gender: 208 males, 276 females

General Outpatient Clinic, Portugal

Prevalence of depression -: 210/484
Cut-off 12
Sensitivity = 0.87
Specificity = 0.73
Watson 2004

Quality assessed: +
GDS-30DSM-IVN = 84, Age over 70 and residing in two Continuing Care Retirement
Communities in US.
Mean age 82

Male – 26%

Prevalence of depression10/78
Major Depression

GDS-30
Standard cut-off ≥ 12
Sensitivity –60% (50, 70)
Specificity –93% (88, 98)
PPV – 55%
NPV – 95%
AUC – 0.88

GDS-30 Alternative cut-offs
Cut-off ≥ 4
Sensitivity – 100%
Specificity - 42%

Cut-off ≥ 5
Sensitivity – 90%
Specificity - 57%

Cut-off ≥ 6
Sensitivity – 80%
Specificity - 68%

Cut-off ≥ 7
Sensitivity – 80%
Specificity - 73%

Cut-off ≥ 8
Sensitivity – 88%
Specificity - 77%

Cut-off ≥ 9
Sensitivity – 80%
Specificity - 85%
ROC analysis – captured 80% of cases

Cut-off ≥ 10
Sensitivity – 60%
Specificity - 88%

Cut-off ≥ 11
Sensitivity – 60%
Specificity - 89%

Cut-off ≥ 12
Sensitivity – 60%
Specificity - 93%

Cut-off ≥ 13
Sensitivity – 60%
Specificity - 97%

Cut-off ≥ 14
Sensitivity – 60%
Specificity - 99%

Cut-off ≥ 16
Sensitivity – 60%
Specificity - 100%

Minor depression

GDS-30
Standard cut-off ≥ 12
Sensitivity – 33% (23, 43)
Specificity – 88% (81, 95)
PPV – 18%
NPV – 95%
AUC – 0.71
Physical health problems
Agrell 1989

Quality assessed: +
GDS-30Psychiatric interviewN = 40 adults attending an outpatient clinic following a stroke.

Mean age – 80 years

Male – 45%

Prevalence of depression:- 17/40
Depression

Recommended cut-off ≥ 10 – GDS-30
Sensitivity – 88%
Specificity – 64%
PPV – 58%
NPV – 88%
Jackson 1993

Quality assessed: +
GDS-30GMSS -AGECATN = 59 hospitalised medially ill older adults.

Mean age – 77.4 years

Male – no reported

Prevalence of depression21/59 (36%)
Depression
Cut-off ≥ 9 – GDS-30
Sensitivity – 100%
Specificity – 55%
PPV – 56%

Cut-off ≥ 10 – GDS-30
Sensitivity – 91%
Specificity – 63%
PPV – 58%

Cut-off ≥ 11 – GDS-30
Sensitivity – 86%
Specificity – 76%
PPV – 67%

Cut-off ≥ 12 – GDS-30
Sensitivity – 81%
Specificity – 74%
PPV – 74%

Cut-off ≥ 13 – GDS-30
Sensitivity – 62%
Specificity – 87%
PPV – 72%
Johnson 1995

Quality assessed: +
GDS-30DSM-III (SCID)N=204 (who received at least one screen and underwent the psychiatric assessment).

GHQ-26: N = 66
HADS N = 93
GDS: N= 120

Mean age (whole sample) = 71 years.

Male: Female = 1.27: 1

Prevalence of depression (whole sample) – 26/204

Prevalence of major depression (whole sample)- 17/204
Any depression

Threshold 13/14
Sensitivity – 84%
Specificity – 50%
PPV – 44%
NPV – 87%

Threshold 10/11
Sensitivity – 84%
Specificity – 66%
PPV – 53%
NPV – 90%

Threshold 11/12
Sensitivity – 74%
Specificity – 70%
PPV – 53%
NPV – 85%
Low 2007

Quality assessment +
GDS-30DSM-IV (SCID-I /NP)N = 119 patients meeting criteria for either acute MI or unstable angina pectoris.

Male = 75%

Mean age = 62.97 (SD 11.61)

Canada, British Columbia

Prevalence of depression7/119
MDD

Cut-off ≥ 10 – GDS
Sensitivity – 100%
Specificity – 79%
PPV – 21%
NPV – 100%

Standard cut-off ≥ 11 – GDS
Sensitivity – 100%
Specificity – 83%
PPV – 25%
NPV – 100%

Cut-off ≥ 12 – GDS
Sensitivity – 100%
Specificity – 88%
PPV – 32%
NPV – 100%

Cut-off ≥ 13 – GDS
Sensitivity – 100%
Specificity – 90%
PPV – 35%
NPV – 100%

Cut-off ≥ 14 – GDS
Sensitivity – 100%
Specificity – 94%
PPV – 50%
NPV – 100%

Cut-off ≥ 14 – GDS
Sensitivity – 67%
Specificity – 94%
PPV – 40%
NPV – 98%

AUC – 0.97

Any Depression

Cut-off ≥ 9 – GDS
Sensitivity – 100%
Specificity – 74%
PPV – 21%
NPV – 100%

Cut-off ≥ 10 – GDS
Sensitivity – 100%
Specificity – 80%
PPV – 25%
NPV – 100%

Standard cut-off ≥ 11 – GDS
Sensitivity – 100%
Specificity – 84%
PPV – 29%
NPV – 100%

Cut-off ≥ 12 – GDS
Sensitivity – 100%
Specificity – 89%
PPV – 37%
NPV – 100%

Cut-off ≥ 13 – GDS
Sensitivity – 100%
Specificity – 91%
PPV – 41%
NPV – 100%

Cut-off ≥ 14 – GDS
Sensitivity – 86%
Specificity – 94%
PPV – 50%
NPV – 99%

AUC – 0.96
Rovner 1997

Quality assessed: +
GDS-30DSM-IVN=70,

Mean Age = 77 years,

Gender: 41 females, 29 males

Prevalence of depression 27/70
Depression

Standard cut-off - GDS
Sensitivity = 63%
Specificity = 77%
Tang 2004B

Quality assessed: +
GDS–30 Chinese versionDSM-IVN= 127 Chinese geriatric stroke patients;

Mean age = 75.7 (SD = 6.2)
Male - 53.5%

Prevalence of depression8/100
Any depression

Optimal cut-off ≥ 7
AUC – 0.90
Sensitivity – 89%
Specificity – 73%
PPV – 37%
NPV – 97%
Ertan 2005

Quality assessed: +
GDS– 30 (Turkish version)DSM-IVN – 109 patients with Parkinson's Disease

Male = 67%

Mean age = 66.5; age range 29-84

Turkey, Istanbul

Prevalence of depression56/109
Depression

Cut-off ≥ 10
Sensitivity – 89%
Specificity – 62%
PPV – 71%
NPV – 84%
Community
Carrete 2001

Quality assessed: +
GDS-30DSM-IV (SCID)N= 169 Mean age = 72 years
gender: 57 males, 112 female

Ambulatory older adults were contacted by telephone, Argentina

Prevalence of depression - 22/169
Cut-off 11
Sensitivity = 0.88
Specificity = 0.84
Costa 2006

Quality assessed: +
GDS-30ICD-10N=126, Older adults, Brazil

Mean age = 81 years,

gender: 36 males, 90 females

Prevalence of depression - 65/126
GDS
Sensitivity = 0.733
Specificity = 0.654
Dunn 1989

Quality assessed: +
GDS-30DSM-III measured used the Depression symptom checklist and the research diagnostic criteria/N = 439 community dwelling older adults attending either an activity centre or dining facility

Mean age – 74 years

Male – % not reported

Prevalence of depression- 36/439
Major depression
Cut-off 11 – GDS 30
False Positive – 53 (18%)
False Negative – 6 (17%)
Sanchez-Garcia 2008

Quality assessed: ++
GDS-30DSM-IVN =534, older adults receiving IMSS, living in Mexico City, 206 individuals randomly selected for a clinical assessment.

Mean age – 71.5 years (SD 7.0years)

Male – 32%

Prevalence of major depression-: 19/206

Prevalence of any depression-: 62/206
Any depression

Standard cut-off GDS
Sensitivity – 53.8% (53.1-54.5)
Specificity – 78.9% (78.4–79.5)
PPV – 60.8% (60.0-61.6)
NPV – 73.7% (73.3-74.1)
Geriatric Depression Scale – 15 item (and Brief GDS)
StudyIdentification toolComparator/ casenessPopulationResults
Consultation
Abas 1998

Quality assessed: +
GDS-15GMS-AGECATN = 164 (82 completed both the screen and the diagnostic interview)

African-Caribbean adults aged over 60 using primary care services/London, UK

Prevalence of depression22/82

Prevalence of depression based on whole sample 20% (95%CI 17-23)
Major depression

Cut-off ≥4
Sensitivity – 89.1%
Specificity - 65.8%

Cut-off ≥5
Sensitivity – 81.5%
Specificity - 81.5%

Cut-off ≥6
Sensitivity – 74.0%
Specificity - 85.5%
Arthur1999

Quality assessed: +
GDS-15ICD-10 based on SCANN = 201

All people aged over 75 in one large GP practice list undergoing a health check.

Leicester, UK

Prevalence of depression- 12/201 – 6%
Depression

Cut-off ≥2
Sensitivity – 100% Specificity - 49.9%
PPV – 11.2%
NPV – 100.0%

Cut-off ≥3
Sensitivity – 100%
Specificity - 71.9%
PPV – 18.4%
NPV – 100.0%

Cut-off ≥4
Sensitivity – 80%
Specificity - 81.6%
PPV – 21.6%
NPV – 98.5%

Cut-off ≥5
Sensitivity – 60.0%
Specificity - 89.2%
PPV – 26.1%
NPV – 97.2%

Cut-off ≥6
Sensitivity – 50.0%
Specificity - 93.7%
PPV – 33.3%
NPV – 96.7%

Cut-off ≥7
Sensitivity – 43.3%
Specificity - 96.0%
PPV – 40.6%
NPV – 96.4%
Blank2004

Quality assessed: +
GDS - 15Diagnostic Interview ScheduleN = 360, participants were recruited from primary care (N=125), general hospitals (N=150) and nursing home (N=85) settings (analysis presented separately for each group). All participants were aged >60 years

Mean age – 77 years

Male = 37%

Prevalence of major depression – 9%

Prevalence of any depression – 16%

Prevalence of major depression in primary care – 11%

Prevalence of major depression in hospital – 8%

Prevalence of major depression in nursing homes – 9%
Major depression

Primary care sample

GDS-15
Cut-off ≥6
Sensitivity – 79% (51-94)
Specificity - 75% (71-77)

AUC – 0.81 (0.67-0.97)

Cut-off ≥9 – recommended
Sensitivity – 71% (45-90)
Specificity - 91% (88-93)

Nursing Home sample

GDS-15
Cut-off ≥6
Sensitivity – 86% (44-99)
Specificity - 82% (78-83)

AUC – 0.87 (0.74- 1.00)

Cut-off ≥7 – recommended
Sensitivity – 86% (44-99)
Specificity - 83% (80-85)

Hospital sample

GDS-15
Cut-off ≥6
Sensitivity – 83% (52-97)
Specificity - 80% (77-81)

AUC – 0.82 (0.68- 0.96)

Cut-off ≥6 – recommended
Sensitivity – 83% (53-97)
Specificity - 80% (77-81)
Cullum 2006

Quality assessed: +
GDS-15ICD-10N = 618 medically ill older adults in hospital settings. Of these 221 completed both the screens and the diagnostic interviews.

Mean age (whole sample) – 80.2years (SD 7.48 years)

Mean age (interview sample) – 80.3 years (SD 7.49 years)

Male (whole sample) – 41%

Male (interview sample) – 40%

Prevalence of depression: - 17.7% (weighted prevalence)
Depression

Cut-off ≥ 5 – GDS-15
Sensitivity – 0.91 (0.71-0.98)
Specificity – 0.63 (0.55-0.71)

Cut-off ≥ 6 – GDS-15
Sensitivity – 0.78 (0.58-0.90)
Specificity – 0.74 (0.66-0.80)

Cut-off ≥ 7 – GDS-15
Sensitivity – 0.74 (0.54-0.87)
Specificity – 0.81 (0.75-0.86)

Cut-off ≥ 8 – GDS-15
Sensitivity – 0.61 (0.43-0.76)
Specificity – 0.86 (0.82-0.89)

Cut-off ≥ 9 – GDS-15
Sensitivity – 0.50 (0.35-0.65)
Specificity – 0.92 (0.88-0.94)

Cut-off ≥ 10 – GDS-15
Sensitivity – 0.39 (0.27-0.52)
Specificity – 0.94 (0.92-0.96)
D'Ath 1994

Quality assessed: +
GDS-15GMSN=194, Age: 74 years, Gender: 126 females, 72 males

Prevalence of depression - 67/194
Depression

Sensitivity 91%
Specificity 72%
Friedman 2005

Quality assessed: +
GDS-15MININ = 960 functionally impaired but cognitively intact older adults participating in a RCT assessing a primary care health intervention. USA

Mean age – 79.3years (SD 7.4 years)

Male – 25.4%

Prevalence of depression: - 124/960 (12.9%)
Depression

Standard Cut-off ≥6
Sensitivity – 81.45%
Specificity – 75.36%

AUC – 0.858 (SE – 0.018)
Hoyl 1999

Quality assessed: +
GDS-15

GDS-5
Clinical evaluation – including MINI, PRIME-MD and psychiatric consultationN=74, frail older adult outpatients. California, USA

Mean age – 74 years

Male – 98%

Prevalence of depression34 / 74 (46%)
Any depression

GDS-15
Sensitivity - 0.94
Specificity - 0.82
PPV – 0.82
NPV – 0.94
AUC – 0.91

GDS-5
Optimal cut off ≥ 2
Sensitivity - 0.97
Specificity - 0.85
PPV – 0.85
NPV – 0.97
AUC – 0.94
Jongenelis 2005

Quality assessed: +
GDS-15DSM-IVN= 333, age = 79 years, gender: 104 males, 229 females

Nursing home, Netherlands

Prevalence of depression -:
74/333
Any depression

Cut-off 5
Sensitivity - 0.81
Specificity - 0.63
Lynes 1997

Quality assessed: +
GDS - 15DSM-III-RN = 130 older adults attending primary care.

Mean age – 71.0 years (SD – 6.8 years)

Male – 41.5%

Prevalence of major depression – 14/130

Prevalence of any depression – 24/130
Major depression

Cut-off 5 GDS-15
Sensitivity = 92%
Specificity = 81%

AUC – 0.935 (0.046)
Marc 2008

Quality assessed: +
GDS-15DSM-IV using SCID and expert consensusN = 526 older adults who were newly admitted to receive home nursing care. Participants with cognitive impairment were excluded from the study. (492 cases used in the analysis due to missing data)

Mean age = 78.3years (SD – 7.5 years)

Male – 34.9%

Prevalence of depression: - 81/526 (15.4%)
Depression

Optimal cut off ≥ 5 – GDS-15
Sensitivity – 71.8%
Specificity – 78.2%

AUC – 0.7933 (SE – 0.0308)

Standard cut off ≥ 5 – GDS-15
Sensitivity – 60.6%
Specificity – 86.2%
Nam Bae 2004

Quality assessed: ++
Short GDS – Korean version (SGDG-K)DSM-III-RN = 154 (91.1% of eligible participants)

Consecutively registered elderly psychiatric patients aged 55+ who visited the Geriatric Psychiatry Clinic in Seoul. People with dementia or any form of cognitive impairment were excluded from the study.

Mean age = 66 years (SD = 6.48yrs)

Male – 35%

Prevalence of depression62/154
Major depression

SGDS-K
Optimal cut-off ≥ 8
Sensitivity = 0.8548
Specificity = 0.6957

Optimal cut-off ≥ 10 (indicated by ROC curve)
Sensitivity = 0.7419
Specificity = 0.8587
Neal 1994

Quality assessed: +
GDS-15DSM (GMS)N=45, Age = 77years, Gender: 18 males, 27 females

Prevalence of depression - 8/45
Depression

Optimal cut-off - GDS-15
Sensitivity 0.67
Specificity 0.80
Pomeroy 2001

Quality assessed: +
GDS - 4

GDS - 15
ICD-10N = 87 patients over the age of 60 admitted to medical rehabilitation wards or attending day rehabilitation facilities; 40% male, mean age 78.4 (SD – 7.7 yrs)

Prevalence of depression17/87
Depressive episode

GDS-4
Optimal cut-off ≥ 1
Sensitivity – 82.4%
Specificity – 67.1%
AUC – 0.80 (0.68, 0.93)
PPV – 37.8%
NPV – 94.0%

GDS-15
Optimal cut-off ≥ 5
Sensitivity – 82.4%
Specificity – 60.0
AUC – 0.82 (0.71, 0.93)
PPV – 33.3%
NPV – 93.3%
Rinaldi 2003

Quality assessed: +
GDS-15

5-item GDS (Hoyl1999) – (GDS-5)
DSM-IVN= 181
Participants were 65yrs and older, with normal cognitive function enrolled from three settings: an acute geriatric ward (33%), a geriatric outpatient clinic (28%) and a nursing home (39%); mean age 79.4 (SD- 7.3yrs)

Prevalence of depression87/181
Any depression

GDS-15
Sensitivity – 0.92 (0.88, 0.96)
Specificity – 0.83 (0.78, 0.88)
PPV – 0.83 (0.78, 0.88)
NPV – 0.92 (0.88, 0.96)
AUC – 0.88

GDS-5
Sensitivity – 0.94 (0.91, 0.98)
Specificity – 0.81 (0.75, 0.87)
PPV – 0.81 (0.75, 0.87)
NPV – 0.94 (0.90, 0.97)
AUC – 0.85
Scheinthal 2001

Quality assessed: ++
GDS-15DSM-IVN=75, Age: 74 years, Gender: 33 males, 42 females

US geriatric medical setting

Prevalence of depression8/75
Cut-off ≥ 7
Sensitivity 1
Specificity 0.79
Van Marwijk 1995

Quality assessed: +
GDS – 15 itemDSM-IIIN=586 age = 65-94 years, gender: 237 males, 349 females

Older people in primary care, Netherlands

Prevalence of depression - 33/586
Any depression

Cut-off <3/3 – GDS-15
Sensitivity – 67%
Specificity – 73%
PPV – 13%
NPV – 97%

Cut-off <2/2+
Sensitivity – 76%
Specificity – 53%
PPV – 9%
NPV – 97%
Physical health problems
Galaria 2000

Quality assessed: +
GDS-15DSM-III-RN = 70 older adults aged over 65, with visual impairments attending a low vision specialist clinic.

Mean age – 77.4 years (SD = 6.6 years)

Male – 41.6%

Prevalence of depression: - 27/70 (38.6%)
Depression

Standard Cut-off ≥ 5
Sensitivity – 0.74
Specificity – 0.72
Haworth 2007

Quality assessed: +
GDS-15DSM-IV (SCID)N=88, Heart failure patients, US

Mean Age = 70 years

Gender: 73 males, 15 females

Prevalence of depression -:
22/88 depression
13/88 MDD
Depression

Cut-off 5 (recommended and optimal)
Sensitivity 81.8%
Specificity 83.3%
PPV 62.1%
NPV 93.2%
Jackson 1993

Quality assessed: +
GDS-15GMSS -AGECATN = 59 hospitalised medially ill older adults.

Mean age – 77.4 years

Male – no reported

Prevalence of depression21/59 (36%)
Depression
Cut-off ≥ 4 – GDS-15
Sensitivity – 100%
Specificity – 50%
PPV – 53%

Cut-off ≥ 5 – GDS-15
Sensitivity – 86%
Specificity – 66%
PPV – 58%

Cut-off ≥ 6 – GDS-15
Sensitivity – 67%
Specificity – 79%
PPV – 64%
Koenig 1992B
(followed on from Koenig 1992 A but used a difference sample in the validation study)

Quality assessed: +
GDS-11DSM-III-RN = 78 males completed GDS-11 and psychiatric interview out of 117 participants who completed the GDS-11 (only those in the first two months of the study had a psychiatric interview).

Participants were all recruited from a neurology unit.

Mean age (of whole 117 sample) = 34.4 years (SD4.7 years)

Male – 100%

Prevalence of depression12/78
Depression

Cut-off ≥ 3 - GDS-11
Sensitivity – 83%
Specificity - 77%
Lee 2008

Quality assessed: +
GDS-15 - Chinese versionDSM-IVN = 253 Stroke patients 1 month after admission to the stroke unit.

Mean age – not reported

Male – 62.8%

Prevalence of depression 116/253
Depression

Cut-off ≥ 5 – GDS-15
Sensitivity – 83.6%
Specificity – 76.6%
PPV – 75.2%
NPV – 84.7%
Tang 2004A

Quality assessed: +
GDS–15 Chinese versionDSM-III-RN = 60 Chinese patients received rehabilitation after stroke

Prevalence of depression 14/60
Any depression

Optimal cut-off ≥ 6
AUC – 0.758
Sensitivity – 64%
Specificity – 83%
PPV – 53%
NPV – 88%
Weintraub 2006

Quality assessed: +
GDS -15DSM-IVN=148 with idiopathic PD receiving specialist care

Mean age = 71 years

MMSE = 27
AUC – 0.92 (0.87, 0.93)

Cut-off 1/2
Sensitivity – 100%
Specificity – 35%
PPV – 30%
NPV – 100%

Cut-off 2/3
Sensitivity – 97%
Specificity – 51%
PPV – 35%
NPV – 98%

Cut-off 3/4
Sensitivity – 91%
Specificity – 71%
PPV – 46%
NPV – 96%

Cut-off 4/5
Sensitivity – 88%
Specificity – 85%
PPV – 61%
NPV – 96%

Cut-off 5/6
Sensitivity – 78%
Specificity – 91%
PPV – 69%
NPV – 93%

Cut-off 6/7
Sensitivity – 66%
Specificity – 97%
PPV – 84%
NPV – 91%

Cut-off 7/8
Sensitivity – 50%
Specificity – 97%
PPV – 84%
NPV – 88%
Community
De Craen 2003

Quality assessed: +
GDS-15DSM-IVN=79, Community dwelling, older adults

Median age = 87 years, gender: 24 males, 55 females

Netherlands
Cut-off 3
True Positive = 7
False Positive = 17
False Negative =1
True Negative =54
Orcos 2007

Unable to quality assess as full translation required -(Detailed English abstract containing information on population and all results)
GDS-15

GDS-5
DSM-IVN= 301, non-selected older community dwelling adults.

Prevalence of depression: - 14.6%
Depression

GDS-15
Sensitivity – 0.818 (0.704-0.932)
Specificity – 0.977 (0.958-0.995)
PPV – 0.857 (0.751-0.963)
NPV – 0.969 (0.948-0.99)

GDS-5
Sensitivity – 0.864 (0.762-0.965)
Specificity – 0.856 (0.813-0.899)
PPV – 0.507 (0.394-0.62)
NPV – 0.973 (0.952-0.994)
Rait 1999

Quality assessed: +
GDS-15DSM-IVN=130,

Mean age = >60 years,

Gender: no information

Prevalence of depression - : 13/130
Depression

Sensitivity - 91%
Specificity - 72%

From: Appendix 20, Case identification included and excluded studies

Cover of Depression in Adults with a Chronic Physical Health Problem
Depression in Adults with a Chronic Physical Health Problem: Treatment and Management.
NICE Clinical Guidelines, No. 91.
National Collaborating Centre for Mental Health (UK).
Leicester (UK): British Psychological Society; 2010.
Copyright © 2010, The British Psychological Society & The Royal College of Psychiatrists.

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Enquiries in this regard should be directed to the British Psychological Society.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.